Benchmarking reports

from sklearn_benchmarks.report import Reporting

reporting = Reporting(config_file_path="config.yml")
reporting.run()

Time report

	hour	min	sec
algo
KNeighborsClassifier	0.0	13.0	39.490230
daal4py_KNeighborsClassifier	0.0	5.0	15.148581
KNeighborsClassifier_kd_tree	0.0	6.0	44.865003
daal4py_KNeighborsClassifier_kd_tree	0.0	1.0	45.613518
KMeans	0.0	12.0	32.549946
daal4py_KMeans	0.0	8.0	24.129848
LogisticRegression	0.0	0.0	4.227819
daal4py_LogisticRegression	0.0	0.0	4.533975
Ridge	0.0	0.0	0.999709
daal4py_Ridge	0.0	0.0	0.716903
total	0.0	48.0	32.353383

KNeighborsClassifier vs daal4py

Results

Plots

KNeighborsClassifier_kd_tree vs daal4py

Results

Plots

KMeans vs daal4py

Results

Plots

LogisticRegression vs daal4py

Results

Plots

Ridge vs daal4py

Results

Plots

import pandas as pd
df_skl = pd.read_csv("results/benchmarking/sklearn_LogisticRegression.csv")
display(df_skl)
df_d4p = pd.read_csv("results/benchmarking/daal4py_LogisticRegression.csv")
display(df_d4p)

	estimator	function	mean	stdev	n_samples_train	n_samples	n_features	hyperparams_digest	dataset_digest	C	...	max_iter	multi_class	n_jobs	penalty	random_state	solver	tol	verbose	warm_start	n_iter
0	LogisticRegression	fit	0.209276	0.005438	1000	1000	1000	f11be2529b0a77bc8fb5ee8de141e6b2	87df8708b64e2831c82c5e60d77c73aa	1.0	...	100	auto	NaN	l2	NaN	lbfgs	0.0001	0	False	[46]
1	LogisticRegression	predict	0.000093	0.000052	1000	1	1000	f11be2529b0a77bc8fb5ee8de141e6b2	87df8708b64e2831c82c5e60d77c73aa	1.0	...	100	auto	NaN	l2	NaN	lbfgs	0.0001	0	False	NaN
2	LogisticRegression	predict	0.000336	0.000079	1000	100	1000	f11be2529b0a77bc8fb5ee8de141e6b2	87df8708b64e2831c82c5e60d77c73aa	1.0	...	100	auto	NaN	l2	NaN	lbfgs	0.0001	0	False	NaN
3	LogisticRegression	fit	0.127952	0.001868	10000	10000	100	f11be2529b0a77bc8fb5ee8de141e6b2	42d2d17798f0471ec3d5da7cbdf7681d	1.0	...	100	auto	NaN	l2	NaN	lbfgs	0.0001	0	False	[20]
4	LogisticRegression	predict	0.000087	0.000052	10000	1	100	f11be2529b0a77bc8fb5ee8de141e6b2	42d2d17798f0471ec3d5da7cbdf7681d	1.0	...	100	auto	NaN	l2	NaN	lbfgs	0.0001	0	False	NaN
5	LogisticRegression	predict	0.000330	0.000071	10000	1000	100	f11be2529b0a77bc8fb5ee8de141e6b2	42d2d17798f0471ec3d5da7cbdf7681d	1.0	...	100	auto	NaN	l2	NaN	lbfgs	0.0001	0	False	NaN

6 rows × 25 columns

	estimator	function	mean	stdev	n_samples_train	n_samples	n_features	hyperparams_digest	dataset_digest	C	...	max_iter	multi_class	n_jobs	penalty	random_state	solver	tol	verbose	warm_start	n_iter
0	LogisticRegression	fit	0.234960	0.022490	1000	1000	1000	f11be2529b0a77bc8fb5ee8de141e6b2	87df8708b64e2831c82c5e60d77c73aa	1.0	...	100	auto	NaN	l2	NaN	lbfgs	0.0001	0	False	[51]
1	LogisticRegression	predict	0.000351	0.000113	1000	1	1000	f11be2529b0a77bc8fb5ee8de141e6b2	87df8708b64e2831c82c5e60d77c73aa	1.0	...	100	auto	NaN	l2	NaN	lbfgs	0.0001	0	False	NaN
2	LogisticRegression	predict	0.000508	0.000111	1000	100	1000	f11be2529b0a77bc8fb5ee8de141e6b2	87df8708b64e2831c82c5e60d77c73aa	1.0	...	100	auto	NaN	l2	NaN	lbfgs	0.0001	0	False	NaN
3	LogisticRegression	fit	0.137598	0.001401	10000	10000	100	f11be2529b0a77bc8fb5ee8de141e6b2	42d2d17798f0471ec3d5da7cbdf7681d	1.0	...	100	auto	NaN	l2	NaN	lbfgs	0.0001	0	False	[21]
4	LogisticRegression	predict	0.000276	0.000105	10000	1	100	f11be2529b0a77bc8fb5ee8de141e6b2	42d2d17798f0471ec3d5da7cbdf7681d	1.0	...	100	auto	NaN	l2	NaN	lbfgs	0.0001	0	False	NaN
5	LogisticRegression	predict	0.000379	0.000103	10000	1000	100	f11be2529b0a77bc8fb5ee8de141e6b2	42d2d17798f0471ec3d5da7cbdf7681d	1.0	...	100	auto	NaN	l2	NaN	lbfgs	0.0001	0	False	NaN

6 rows × 25 columns